The objective of this study was to determine whether pretest probability assessments permit more selective testing of chest pain patients with technetium-99m sestamibi scanning. Pretest probabilities of cardiac ischemia were measured both objectively (Acute Cardiac Ischemia Time-Insensitive Predictive Instrument [ACI-TIPI]) and subjectively (physician's estimate of the probability of unstable angina). Two groups were defined: patients whose postsestamibi scan led to a "downgrade" of the intensity of monitoring and those that resulted in no change in monitoring intensity. Sixty-five patients met study criteria; 25 had a disposition downgrade and 40 had no change. Pretest ACI-TIPI scores were similar in the two groups (29% +/- 18% versus 27% +/- 11%, mean +/- standard deviation; P = .95) as were the physician's assessment of unstable angina (39% +/- 22% versus 40% +/- 24%; P = .75). Objective or subjective pretest probabilities are not significantly different in patients who are likely to have their disposition altered by sestamibi scanning.